Unemployment Insurance and Takeovers* Lixiong Guo† University of Alabama Jing Kong‡ Michigan State University Ronald W. Masulis§ University of New South Wales December 11, 2019 Abstract We examine whether unemployment insurance (UI) mitigates shareholder-employee conflicts of interest and improves takeover market efficiency. Exploiting within state changes in UI benefits, we find that increased target state UI benefits raise acquisition likelihoods, expected deal synergies, and gains to both acquirer and target shareholders. UI effects on acquisition likelihood are stronger in highly unionized industries, in firms with concentrated employee ownership, and in firms with employee friendly boards. The passage of directors’ duties laws increases the UI effect on acquisition likelihood. Lastly, increased acquirer state UI benefits raise acquirer returns and likelihood of acquirers making within-industry acquisitions. JEL Classification: G32, G34, G38, J65 Keywords: Unemployment Insurance, Labor, Takeovers, Mergers and Acquisitions, Synergy, Stakeholders, Unemployment Risk. * We are grateful to Shai Bernstein, Oleg Chuprinin, Andrey Golubov (discussant), Vidhan Goyal, Chang-Mo Kang, Hao Liang (discussant), Shawn Mobbs, Hernan Ortiz-Molina (discussant), Peter Swan, Li Yang, Xianming Zhou, and seminar participants at UNSW Sydney, Deakin University, University of Alabama, and conference participants at the 2017 NFA Conference in Halifax, 2017 FMA Conference in Boston, 2017 AsianFA Conference in Seoul, and 2017 FIRN Corporate Conference in Adelaide for helpful discussions and valuable comments. All errors are our own. † Lixiong Guo is with the University of Alabama. Address: Culverhouse College of Business, University of Alabama, 232 Alston Hall, 361 Stadium Drive, Tuscaloosa, Alabama, 35487-0224. Electronic Mail: [email protected]. ‡ Jing Kong is with the Michigan State University. Address: Eli Broad College of Business, Michigan State University. Electronic Mail: [email protected]. § Ronald Masulis is with the University of New South Wales. Address: UNSW Business School, UNSW, Sydney, NSW, 2052, Australia. Electronic Mail: [email protected]. 1. Introduction Corporate takeovers play a major role in allocating resources and improving economic efficiency. Yet conflicts of interests among various stakeholders of both acquirers and targets often prevent shareholders from realizing the full economic potential of takeovers. The manager- shareholder conflicts of interest have received substantial attention in the literature. Yet, takeovers not only jeopardize target manager jobs, but also target and acquirer employee jobs. The greater likelihood of employees losing jobs to post-merger workforce restructurings creates shareholder- employee conflicts of interest. Recent studies examine how such conflicts are exacerbated by employee legal rights and employee protection laws.5 But, few studies examine alternative labor market mechanisms that could mitigate these conflicts. We help fill this void by closely examining unemployment insurance (UI) and whether it can serve as an effective mechanism to lower takeover market frictions. In most developed countries, employees who lose their jobs in corporate restructuring are eligible for temporary income provided by public unemployment insurance systems. In the U.S., ninety-seven percent of wage and salary workers are covered by state UI programs, which represents one of the largest government transfer programs in the US. Prior studies find that UI can significantly lower worker unemployment risk and soften the negative income shocks of unexpected unemployment. For example, Hsu, Matsa and Melzer (2018) estimate that an added $3600 in maximum total UI benefits under regular UI programs reduce mortgage delinquencies by 114 basis points, and UI expansion during the 2007-2009 recession prevented more than 1.3 5 Using U.S. acquisitions, John, Knyazeva and Knyazeva (2015) find that acquirers from states with weak labor rights, defined as states with right-to-work laws, experience higher deal announcement returns and higher combined acquirer and target returns. Tian and Wang (2016) find that U.S. firms where employees narrowly vote for union representation receive fewer takeover bids than those that narrowly vote against, and they receive lower takeover premiums when they are acquired. Using data in 21 OECD countries, Dessaint, Golubov and Volpin (2017) find that increased legal protection of employment in target countries reduces both deal activity and deal synergies. 2 million additional foreclosures. Hence, for many unemployed workers, a relatively small change in UI compensation can have a large impact on their lives. Not surprisingly, a number of studies find that UI compensation has economically meaningful effects on worker behavior and major firm policies (Meyer, 1990 and 1995; Gruber, 1997; Engen and Gruber, 2001; Light and Omori, 2004; Gormley, Liu and Zhou, 2010; Agrawal and Matsa, 2013; Ellul, Wang, and Zhang, 2015; Ellul, Pagano and Schivardi, 2018).6 This body of evidence suggests that UI compensation can reduce employee opposition to takeovers and post-merger workforce restructuring, and consequently can improve the efficiency of the market for corporate control. In the U.S., unemployment insurance is provided by a federal-state system in which the basic framework is the same nationwide, but states vary substantially in important program parameters such as the weekly benefit amount. States also exhibit substantial differences in their adjustment of UI benefits over time. Due to uncertain legislative approval of changes in UI laws, the timing of UI benefit changes is often exogenous to underlying economic conditions (Agrawal and Matsa, 2013; Hsu, Matsa and Melzer, 2018). In this study, we exploit these statutory changes within individual states to establish a causal effect of UI benefits on takeover probabilities and shareholder gains. Our sample consists of all U.S. public firms in the CRSP-Compustat merged database over the period 1990-2015, excluding heavily regulated utilities and financial services firms. Following Agrawal and Matsa (2013), we use the product of the maximum weekly benefit amount and the maximum duration of the benefits to measure the overall level of UI benefits in a state (Ffor brevity, 6 Prior studies find that unemployment insurance affects a worker’s unemployment duration (Meyer, 1990 and 1995), consumption smoothing (Gruber, 1997), job quits (Light and Omori, 2004), individuals’ precautionary savings (Engen and Gruber, 2001), stock market participation (Gormley, Liu and Zhou, 2010), a firm’s capital structure (Agrawal and Matsa, 2013) and executive compensation policies (Ellul, Wang, and Zhang, 2015), and a family firm’s provision for employment stability to their workers (Ellul, Pagano and Schivardi, 2018).. 3 UI level hereafter). We focus on target firms because post-merger workforce reductions typically fall much more heavily on target employees.7 In support of our main hypothesis that UI benefits reduce shareholder-employee conflicts of interest at target firms, we find that a 10% increase in UI level raises the likelihood of a firm headquartered in that state being acquired by 94 basis points per annum, which represents a 28% increase over the unconditional takeover likelihood of 3.3%. The effect is statistically significant at the 1% level. Although the uncertain legislative approval of UI law changes suggests that this effect is causal, we conduct explicit identification tests to further substantiate this causal interpretation. We find that the relationship cannot be explained by three major alternative explanations, namely unobserved regional economic shocks, unobserved industry shocks or reverse causality. We hypothesize that three channels are at work, acting either independently or jointly to contribute to this relationship. First, in some target firms, employees have strong legal rights or own large equity positions, which provides them a formal channel to directly influence firm takeover outcomes. Prominent examples are coverage by a collective bargaining agreement and large employee holdings of the firm’s stock through employee stock ownership plans (ESOPs), 401(k) plans and profit-sharing plans, etc. Under the National Labor Relations Act (NLRA), an employer has certain obligations to bargain with a unionized workforce over the effects of a proposed M&A deal or the decision to sell the firm, which gives employees a voice in takeover decisions. In firms where employees as a group hold a large equity block, their votes on a merger proposal or their decisions to tender their shares to the acquirer can be pivotal to the success of a takeover attempt. In either case, if employee opposition to takeovers varies with 7 Another reason for focusing on target firms is that existing evidence on whether strong target labor rights have a significant effect on takeover outcomes in U.S. acquisitions is mixed. By focusing on target state UI benefits, we can see whether properly controlling for UI benefit levels can help to resolve this puzzle. Nonetheless, for the purpose of providing comprehensive evidence on how UI affects takeovers, we also examine how acquirer-state UI benefits reduce acquirer shareholder-employee conflicts of interest 4 UI level, then changes in UI level can directly affect the takeover probabilities of these firms. In support of this proposition, we find that UI has a stronger effect on takeover likelihoods when target firms are in highly